Résumé
Ironically, wetland raised-field (RF) agriculture has attracted most research attention in the Neotropics, where, with few exceptions, it disappeared 500 years ago. Huge areas of vestiges of pre-Columbian RFs have been studied since the 1960s by geographers and archaeologists in numerous sites ranging from Mexico and Guatemala to the Bolivian llanos. Although two pioneers in the study of pre-Columbian RFs, William Denevan and Billy Turner, drew attention 45 years ago to the widespread practice of RF agriculture in the Old World today, few have followed their lead to develop more detailed studies of these systems. Particularly widespread in Africa, RF agriculture remains woefully understudied. Furthermore, recent work has revealed the existence of large expanses of archaeological RFs in wetlands of west-central Africa. Interdisciplinary studies combining archaeology, ethnoarchaeology, ethnobiology and agronomy hold the promise of exploring the connections between the present and the past. Such studies also offer the prospect of exploring parallels and contrasts in RF agriculture across continents, thereby fostering synthesis and the search for generalizations. We will present data on two recently studied examples of present-day wetland RF agriculture, in the cuvette centrale of the Congo Basin and in the Bangweulu wetlands in Zambia. These systems present similarities but also strong contrasts, permitting exploration of the diversity within RF agriculture, not only today but probably in the past as well. Presenting the first data on functioning of soils under RF agriculture, we will situate these systems within the general context of low-external-input agriculture in grasslands, open woodlands, and other non-forested environments, where low plant biomass limits the potential for ash from slash-and-burn as nutrient amendment. These first studies of RF agriculture in Africa today open numerous questions about past systems, both in Africa and the Neotropics. We will outline some of these.